[OpenAFS-devel] "fakestat" side effect

[Alf Wachsmann]

On Wed, 4 Sep 2002, Nickolai Zeldovich wrote:
> > Would it be better (and possible) to fake the stat information only
> > on mount points for volumes from foreign cells?
>
> It's certainly possible to do this.  Actually I think it would be
> more consistent to, say, only do fakestat for mountpoints that
> cross cells.  I'm not sure which behavior is really better, though.
> Suppose one of the fileservers in your cell is down, and you try
> to browse /afs/@cell/user.  Do you want your client to hang, even
> though you aren't about to access a user volume on the down server?

I like to operate under the principle of "the least surprise" for users.

Having the listing of /afs/@cell/user hang while a local server is down
is what users would expect.

But I can easily anticipate a lot of support questions from users seeing
a weird date on an internal volume mount point.

Perhaps it's time for yet another parameter for "afsd" to control this
behavior ("-fakestat cross" vs. "-fakestat all")?

-- Alf.

-----------------------------------------------------------------------
  Alf Wachsmann                       | e-mail: alfw@slac.stanford.edu
  SLAC Computing Service              | Phone:  +1-650-926-4802
  2575 Sand Hill Road, M/S 97         | FAX:    +1-650-926-3329
  Menlo Park, CA 94025, USA           | Office: Bldg. 50/323
-----------------------------------------------------------------------
                http://www.slac.stanford.edu/~alfw (PGP)
-----------------------------------------------------------------------